Google's AI-powered defense suspended 39 million advertiser accounts

Google has released its 2024 Ads Safety Report, revealing dramatic improvements in their ability to prevent fraudulent advertisers from displaying ads. The report, announced just four days ago on April 16, 2025, details how the technology giant has harnessed artificial intelligence to strengthen its defensive capabilities against bad actors attempting to exploit its advertising platforms.

The numbers are staggering. According to the report, Google suspended over 39.2 million advertiser accounts in 2024, marking a 208% increase compared to the 12.7 million accounts suspended in 2023. This massive surge in enforcement actions demonstrates the company's intensified focus on preventing fraudulent activity before it reaches consumers.

Alex Rodriguez, General Manager of Ads Safety at Google, highlighted the central role of AI in this enforcement evolution. "In 2024, we launched over 50 enhancements to our LLMs, which enabled more efficient and precise enforcement at scale," Rodriguez stated in the announcement. These Large Language Models have fundamentally transformed Google's approach to ad safety, shifting from reactive removals to proactive prevention.

The technical capabilities of these AI systems represent a significant advancement over previous technologies. While traditional machine learning models required extensive training with vast datasets of violative content, Google's latest LLMs operate with remarkable efficiency, needing only a fraction of that information to identify emerging threats.

These AI tools accelerate complex investigations by uncovering networks of bad actors and repeat offenders through analysis of subtle patterns. According to the report, the systems identify suspicious signals like business impersonation and illegitimate payment information during account setup, preventing potentially harmful ads from ever being displayed.

Impressive enforcement statistics

The scale of Google's enforcement actions in 2024 is substantial:

• Removed over 5.1 billion advertisements
• Restricted over 9.1 billion advertisements
• Suspended over 39.2 million advertiser accounts
• Blocked or restricted ads on 1.3 billion publisher pages
• Took broader site-level enforcement action on over 220,000 publisher sites
• Made over 30 updates to its ads and publisher policies

When comparing year-over-year numbers, the most dramatic increase is in account suspensions. While the 5.1 billion ads removed in 2024 represents a slight decrease from the 5.5 billion removed in 2023, the tripling of account suspensions reveals Google's strategic shift toward stopping bad actors earlier in the process.

The effectiveness of this approach is evident in the publisher policy enforcement area, where Google reports that AI-powered models contributed to 97% of detection and enforcement actions on problematic pages. This technology accelerated site reviews, enabling legitimate publishers to monetize content more quickly while preventing ads from appearing on policy-violating pages.

Fighting evolving scam tactics

The report addresses how Google has adapted to combat increasingly sophisticated scam techniques, particularly the rise of AI-generated public figure impersonation ads. These scams use deepfake technology to create convincing but fraudulent endorsements from celebrities and public figures.

In response to this emerging threat, Google assembled a specialized team of over 100 experts to analyze patterns and develop countermeasures. The company updated its Misrepresentation policy to enable rapid suspension of accounts promoting such scams, leading to the permanent suspension of more than 700,000 offending advertiser accounts.

This focused effort yielded impressive results, with Rodriguez stating that it led to "a 90% drop in reports of this kind of scam ad last year." The company blocked or removed over 415 million ads and suspended more than 5 million accounts for violating ads policies most closely associated with various types of scams.

Google's report also details specific policy enforcement categories. The company took action against ads violating policies related to:

• Abusing the ad network (793.1 million ads)
• Trademark issues (503.1 million ads)
• Personalized ads (491.3 million ads)
• Legal requirements (280.3 million ads)
• Financial services (193.7 million ads)
• Misrepresentation (146.9 million ads)

Election integrity focus

With 2024 being a major election year globally, Google emphasized its work on election integrity. The company expanded its identity verification and transparency requirements for election advertisers to additional countries, including mandatory "paid for by" disclosures and a public transparency report.

Google verified more than 8,900 new election advertisers during the year and removed 10.7 million election ads from unverified accounts. The company also highlighted that it was the first major platform to launch disclosure requirements for AI-generated content in election ads, building on its existing transparency efforts.

Cross-industry collaboration

Recognizing that combating scams requires industry-wide coordination, Google co-founded the Global Signal Exchange with the Global Anti-Scam Alliance to enhance information sharing across platforms. This initiative reflects the growing recognition that sophisticated scammers operate across multiple platforms and require a unified defensive response.

"The ad safety landscape is constantly changing, reshaped by technological breakthroughs like advancements in AI, emerging abuse tactics, and global events, demanding continuous agility from our industry," the report states, emphasizing the ongoing nature of this challenge.

Implications for marketers

For the marketing community, these developments highlight the increasing importance of maintaining legitimate advertising practices. The dramatic increase in account suspensions indicates that Google is taking a more aggressive stance against potentially fraudulent advertisers, with AI systems flagging suspicious behavior earlier in the process.

Marketers should ensure they comply with Google's expanding policies and verification requirements, particularly in sensitive categories like financial services, healthcare, and election advertising. The report's emphasis on advertiser identity verification, which now covers more than 200 countries and territories, underscores the growing importance of transparency in digital advertising.

Additionally, the rising sophistication of AI-powered enforcement means that attempts to circumvent policies are increasingly likely to be detected and penalized. For legitimate advertisers, however, these same systems may enable faster approval processes and improved protection from unfair competition by scammers and policy violators.

Timeline of Google's Ads Safety enforcement

• March 27, 2024: Google releases its 2023 Ads Safety Report, highlighting early impacts of generative AI on enforcement efforts
• November 2023: Google launches the Limited Ads Serving policy to limit the reach of advertisers without established track records
• Late 2023-Early 2024: Google identifies and responds to the rise of AI-generated public figure impersonation scams
• 2024: Google implements over 50 enhancements to its LLM-based enforcement systems
• 2024: Google suspends 39.2 million advertiser accounts, up 208% from 12.7 million in 2023
• April 16, 2025: Google announces its 2024 Ads Safety Report, documenting the impact of its enhanced AI enforcement capabilities